In some sense, just about everything we do in life is toxic. There is a probability that just by crossing the street I will be hit by a car. Physical exertion and exercise can precede the onset of a heart attack or stroke. Vegetables, meats and just about every food source on the planet can potentially contain copious amounts of carcinogens. Sunlight releases high levels of toxic rays that damage our cells. Frankly, I find the world to be a pretty scary place.
But let’s clarify that for a moment. Toxicity is typically measured by dose, and in this case the “copious amounts of carcinogens” in everyday foods tend to be rather low on the scale of what our bodies can handle. Similarly, the probabilities of major injury after completing daily activities like taking a walk across the street or jog down the lane are pretty low. Part of the reason why we do so well amid this dangerous world is that our bodies have evolved strategies to defend us. We have well-developed sensory and motor systems to guide us across the street just as we also have highly efficient biochemical machinery to repair DNA damage caused by sunlight or by the metabolism of chemicals like food or alcohol.
In this context, the fact that the DNA in our bodies is constantly undergoing damage is less of a concern that it might immediately appear. Nonetheless, there’s no reason to be slack. It is currently believed that the slow accumulation of damage to DNA (among other things) plays a critically important role in aging and age-related disease. Thus, the causes and consequences of DNA damage in the brain are definitely worth studying.
A recent study published in Nature Neuroscience adds one more to the list of potential harms: thinking can damage your DNA. Actually, it’s not so much “thinking” or doing mental tasks as it is increased “brain activity” in specific regions of the brain (keep in mind that your brain is very active even when you rest or sleep).
The researchers, operating out of the Gladstone Institute of Neurological Disease in California, determined that common experiences such as exploring a novel environment or mild brain activity could promote rapid increases in DNA damage within the brains of mice. For example, the researchers found that when mice were allowed to explore a novel environment, tissue extracted from parietal cortex and the dentate gyrus of the hippocampus (which is often activated during novel experiences) had 2-3 fold higher levels of a marker of DNA repair (called γH2A.X) than mice which had remained in their home cage. Mice which were returned to their cages for 24 hours after the novel experience had normal levels of the marker, suggesting that the damaged DNA had been repaired. This effect mimicked the increase γH2A.X found in Alzheimer Disease model mice expressing neurotoxic levels of beta amyloid.
Additional experiments further demonstrated that activation of visual cortex by presenting mice with a bar grating to one side of the visual field only produced increased levels of γH2A.X in the contralateral (activated) region of visual cortex. Similarly, lateralized (one-sided) optogenetic stimulation of the mouse striatum produced selective increases in γH2A.X in the activated side (selective activation can be easily confirmed because side-specific activation of some striatal circuits will cause mice to rotate ~60 degrees when walking). Most notable, the researchers found that cells in multiple regions of the brain including dentate gyrus of the hippocampus, parietal and visual cortex and striatum all had higher levels of a marker of DNA damage
A final line of evidence strongly supported the role of normal brain activity in producing DNA damage. The authors had groups of mice either stay in their home cages or enter a novel environment (n = 6-9 per group). Immediately afterwards, dentate gyrus tissue was extracted and cell nuclei were run on a neutral pH single-cell electrophoresis procedure called the “comet assay”. Using this assay, nuclear DNA generally remains clustered together in a little ball, but when DNA has been damaged, fragments of double-stranded DNA (dsDNA) can break off and escape the nucleus. These fragments appear to form a comet-like smear behind the clustered nuclear DNA (see right). The authors found that the percentage of dentate gyrus nuclei with comet tails increased from ~15% in the control group to ~35% in the novel exploration group. This finding mimicked the results found when neurons were exposed to low-dose gamma irradiation, a treatment known to cause dsDNA damage.
So what’s the story, does normal brain activity cause toxic DNA damage?
For starters, DNA breaks are not uncommon. Every time a cell divides it must duplicate its DNA, a process which requires the DNA double-helix to be unwound for the replication machinery to bind to and copy each DNA nucleotide. Before this unwinding process can occur, a break must be made in one strand of the DNA. Because these single-stranded breaks must occur so frequently, every cell in the body has machinery called “ligases” to repair them. Other forms of damage can occur which only impact one of the two strands of DNA. For example, ionizing or ultraviolet radiation or various oxidative species produced during metabolism can cause nucleic acids to exchange in incorrect based pair into the DNA double helix. However, if this damage is restricted to only one sidestrand of the double helix the complementary strand can be used as a template to replace the original base.
In contrast, when damage occurs to both strands of DNA in the same location, different mechanisms of repair may be required. For example, if the phosphodiester bonds that connect each DNA nucleotide to its neighbor are broken on both strands at once, this double-stranded break cannot be fixed so easily. There are two key mechanisms for repairing double-stranded breaks. In dividing cells, once S-phase is complete there are two identical copies of each chromosome (called sister chromatids), which provide an exact template for repairing double-stranded DNA in a process called homologous recombination. However, most cells in the adult brain have permanently ceased dividing (although note that the dentate gyrus, is one of the few places were cells do continue to divide), which means that double-stranded breaks must be repaired through an alternate mechanism, called non-homologous end joining (NHEJ). In this instance, there are no identical sister chromatids to copy, which means that broken ends of DNA must be attached directly, which can frequently lead to errors (i.e. when the ends are not compatible).
See a video illustration of NHEJ, below:
DNA repair is critically important because cells with aberrant DNA can fail to function appropriately if protein or RNA products start to work incorrectly. In more extreme cases, these cells may be targeted for cell death or apoptosis or (much more rarely) may become cancerous and start dividing uncontrollably. This could be a major problem in the adult brain because few cells in this structure are replenished by cell division, which means that cells which die due to extensive DNA damage may be gone forever. Given that the human brain appears to function more-or-less appropriately for many decades, it is probably safe to assume that DNA damage is not accumulating at a rapid rate across vast areas of the average adult brain (although this may change in aging). In part this is because of DNA repair machinery can fix many of these breaks–although, as I note above, if the process by which double-stranded breaks are repaired in the brain is highly error prone, then DNA repair may not be sufficient to contain the damage. For this reason, it’s generally been assumed that double-stranded breaks are relatively rare in the adult brain, apart from pathological situations like tumour development or age-related decline and neurodegeneration.
If this most recent study is correct, then our brains may be producing large levels of double-stranded DNA breaks literally every day, just by undergoing normal function.
So either the DNA repair machinery is performing unexpectedly well (very few errors), or the results of this study are worth another look. Skeptical about the results myself, I had a quick look at the literature and some of the scholarly comments accompanying the paper itself.
What pops out as most apparent is that in virtually all of the experiments, the authors didn’t actually measure DNA damage directly. Instead, they measured the modification of γH2A.X, a histone protein which is phosphorylated before binding directly to regions of DNA damage. Cancer researchers have noted that increases in γH2A.X levels correspond closely to increases in DNA damage, suggesting that it is a good marker of DNA damage. On the other hand, there are some studies which indicate that this histone protein is modified in the absence of DNA damage. If this is the case, and this marker indicates not only DNA repair but also epigenetic regulation, then the results of this study would be consistent with gene regulation induced by brain activity, rather than DNA damage.
The one direction demonstration that brain activity induces DNA damage comes from the “comet” assay I mention above. If the results of this assay are inconclusive, then this latter regulatory interpretation of the above results is much more plausible. Writing in the accompanying commentary to this paper, Karl Herrup, Jianmin Chen and Jiali Li note their surprise:
“And yet, we find ourselves asking, “How can this possibly be?” Do the neurons of our brain really do serious damage to their genome every time we execute a mental task? If 2 h of thinking is enough to trigger DSBs in even a small percentage of our nerve cells, then each cell must put its genome in jeopardy many times over the course of a year.”
They argue that the assay might cause the DNA damage that is seen in the “tail” of the comet. That is, if novel exploration by mice causes changes in gene expression and DNA conformation, then this DNA may be more susceptible to mechanical damage caused by the assay itself.
In support of this perspective, I will note that according to the results of this study, approximately 15% of cell nuclei extracted from the dentate gyrus of control animals had comet tails. While I am not a cell biology expert, I noted that a recent Nature Protocols paper indicates that the specific version of the comet assay these authors used (neutral pH gel, see bottom image to right) appears to produce short comet tails in control nuclei, even in the absence of any DNA-damaging gamma irradiation (0 Gy). This method was chosen in preference to the above “alkali” method because it allows definitive demonstration of double-stranded (rather than single-stranded) DNA breaks. However, if it increases the background level of DNA damage, then it may suggest a potential confounding interpretation. Also, DNA damage using this assay is generally measured by measuring the comet tail moment (or the length of the tail * the proportion of DNA in tail), whereas in the present study the researchers only reported a difference in the proportion of nuclei with a tail. The tail lengths and proportion of DNA in the tail (tail moment) were unchanged in the novel exploration relative to the control group. Thus, it might be possible that the comet assay indicates greater susceptibility to mechanical damage (background noise) after brain activity, rather than dsDNA damage itself.
Regardless, Herrup et al suggest that these two distinct interpretations (brain activity induces epigenetic changes vs brain activity induces DNA damage) are easily testable:
“If DNA damage is occurring during normal neuronal activity, then we should see other components of the DSB repair pathway, such as activation of ATM, a phosphatidylinositol-3-OH kinase, and its downstream target proteins, the tumor suppressor p53 and the checkpoint kinase Chk2. But whether the story is one of DNA breaks or fragile sites induced by modification, the findings should give pause to neuroscientists in nearly every discipline. They are solid findings suggesting that there are new and possibly unexplored genetic or epigenetic regulatory mechanisms used by nerve cells during normal nervous system function.”
Like always with new research, It sounds as though this preliminary study will be just the beginning of new insights into how the brain functions.
Suberbielle E, Sanchez PE, Kravitz AV, Wang X, Ho K, Eilertson K, Devidze N, Kreitzer AC, & Mucke L (2013). Physiologic brain activity causes DNA double-strand breaks in neurons, with exacerbation by amyloid-β. Nature neuroscience, 16 (5), 613-21 PMID: 23525040